Photographing a black cat is hard for the same reason rendering one is: most of the frame is fur with almost no visible detail in it. A photo where the cat is a silhouette gives the model very little to work with, and the result is usually a dark shape that claps rather than your cat that claps.
The fix is in the upload, not in the prompt. Look for a photo where light falls on the face - daylight, a window, or a flash that brings out the eyes - and where the fur has some structure: the sheen along the back, the shape of the ears, a visible collar. Those details are what the model holds on to.
Everything else is identical to the standard workflow: two photos, one click, ten seconds of vertical video delivered as a silent MP4. Black cats are simply the case where the source photo matters most, which is why they get their own page.